DISCRIMINATION OF SIZE AND FORM IN THE RAT 313 



Whenever any such reactions were discovered, immediate 

 attempts were made to eliminate them, and control series were 

 given to determine their importance in the total reaction. In 

 general, it was only in the more difficult problems that the 

 animals fell back upon these secondary criteria. The sense of 

 smell never seemed to play any part in the reaction. 



As a rule, each animal was given 20 trials a day, but this 

 number could not be adhered to strictly. With the type of 

 apparatus used, 50 per cent of error indicates that the animals 

 are not discriminating, that their movements are due to chance, 

 and it was found during succeeding experiments that a much 

 lower percentage of error, not more than 25, upon several con- 

 secutive days was necessary to give conclusive proof of dis- 

 crimination. An animal might make 90 per cent of correct 

 choices in a series of 20 trials, but such a high average, if due 

 to chance, never persisted through two consecutive series. The 

 percentage of error is not always a safe measure of the animal's 

 ability to discriminate between the stimuli. It does not take 

 into account the fluctuations of attention or the difficulties of 

 associative and learning processes. For this reason the behavior 

 of the rats in the discrimination compartment has been given 

 special emphasis. 



ANIMALS 



In the experimental work albino rats alone were used. Trapped 

 specimens of Mus rattus proved to be too wild for experimental 

 purposes, and domesticated races of pigmented rats have shown 

 no marked superiority to the albinos in visual discrimination, 

 so no extensive comparison between pigmented and non-pig- 

 mented races was undertaken. 



The rats used differed considerably in age and in behavior 

 at the beginning of the experiments. 



No. 1. Female, 11 weeks old when given this problem; vigor- 

 ous, active, and stable. 



No. 2. Female, from the same litter as No. 1; very large 

 and inactive. 



No. 3. Female, from the same litter as No. 1 ; stunted in 

 growth, very excitable, and easily frightened. 



No. 4. Female, about two months old; large and active. 



No. 5. Female, from same litter as No. 4; ill during the 

 first part of the experiment, small but very active. 



